Abstract Pyridoxine 5'-phosphate oxidase (PNPO) is a rate-limiting enzyme in converting inactive forms of Vitamin B6 (VB6) in diet, including pyridoxine and pyridoxamine, to the only active form, pyridoxal 5'-phosphate (PLP). PLP is a cofactor required for the syntheses of dopamine, serotonin and GABA in the brain. In humans, PNPO deficiency is known to cause neonatal epileptic encephalopathy (NEE). Mutations in PNPO have been increasingly reported in NEE patients. Recent studies also identify PNPO as a contributor to early-onset epilepsies and one of the16 epilepsy genes involved in the common epilepsies. However, due to the lack of animal models, we know little about the developmental or adult functional impact of PNPO deficiency at systems, circuit or cellular level (e.g. involvement of GABA, dopamine or serotonin synthesis) under in vivo conditions. We know little about how mild PNPO deficiency interacts with other genetic defects or environmental factors (e.g. VB6 in diet) to cause seizures or other conditions. We have identified the Drosophila homolog of PNPO and identified a Drosophila mutant (sgll95 flies) with partial PNPO deficiency. Due to low PNPO activity, they are sensitive to dietary VB6 deficiency. We have since generated global knock-down as well as knock-in models in which the endogenous wild-type (WT) fly PNPO was replaced by human mutant PNPO found in patients. Viability during development, lifespan and seizure phenotype of these flies depend on the specific genetic manipulation as well as availability of VB6 in diet. We have also found that PNPO deficiency exacerbated other epileptic mutant alleles in flies with significant synergistic interactions. In Aim 1, we will define specific developmental stages in fly models in which PNPO deficiency leads to lethality and seizures. In Aim 2, we will define brain specific cell types involved in PNPO- deficiency-induced lethality and seizures in fly and mouse models. We will test gene-gene interactions (e.g., PNPO and other known epilepsy genes). In Aim 3, we will generate and characterize fly and mouse models that carry human PNPO mutations. 1